Docs/visualization training

docs/source/actor_visualization.rst

@@ -29,7 +29,7 @@ Available Settings
 
 * Optional
 
-  - ***port***: (int) The port on which the visualization server will run. Defaults to `5006`.
+  - ***port***: (int) The port on which the visualization server will run. Defaults to `0`, indicating a random available port.
   - ***throttle_interval***: (float) The minimum time in seconds between plot updates. 
     This can be used to prevent the visualization from slowing down if data arrives very quickly. Defaults to `0.1`.
   - ***keep_alive***: (bool) If `True`, the visualization server will remain active after 

docs/source/training.rst

@@ -40,8 +40,8 @@ Exercise 1a: Setting Up Your First Visualization
       We will start by running the visualization actor for a simple example configuration. 
       First, create a yMMSL configuration file that sets up a simple visualization pipeline with:
 
-      1. A source actor that sends the equilibrium IDS
-      2. A visualization actor that receives and plots the data.
+      1. A :ref:`source actor <actor_sink_source>` that sends the equilibrium IDS
+      2. A :ref:`visualization actor <actor_visualization>` that receives and plots the data.
 
       Use the following settings in the yMMSL:
 
@@ -60,7 +60,8 @@ Exercise 1a: Setting Up Your First Visualization
       What do you see in your browser?
 
       .. hint::
-         Take a look at the example yMMSL file in ``imas_muscle3/visualization/examples/simple_1d_plot/simple_1d_plot.ymmsl``. If you want detailed information about the visualization actor, take a look
+         There are premade examples available that you use can use available, located in this
+         location: ``imas_muscle3/visualization/examples``. For this specific exercise, take a look at the example yMMSL file in ``imas_muscle3/visualization/examples/simple_1d_plot/simple_1d_plot.ymmsl``. If you want detailed information about the visualization actor, take a look
          at the :ref:`documentation <actor_visualization>`.
 
    .. md-tab-item:: Solution
@@ -99,7 +100,7 @@ Exercise 1a: Setting Up Your First Visualization
            visualization_component:
              threads: 1
 
-      When you launch the muscle_manger, the browser should open, and you will see the
+      When you launch the muscle_manager, the browser should open, and you will see the
       plasma current plotted over time, updating in real-time as the new time slices are 
       received by the visualization actor.
 
@@ -137,14 +138,15 @@ Exercise 1b: Understanding the Basic Structure
       The ``State`` class **must** implement the ``extract(self, ids)`` method.
       The ``extract`` method for this example case:
 
-      - Checks if the incoming IDS is an equilibrium IDS.
+      - handles every IDS that is received on the S port, one at a time. So first it checks if the incoming IDS is an equilibrium IDS.
       - Extracts the plasma current of the time slice (``ids.time_slice[0].global_quantities.ip``) and 
         its corresponding time value (``ids.time[0]``), and stores it in an Xarray dataset.
       - Either stores the first Xarray dataset entry in ``self.data`` or appends it to the existing Xarray dataset.
 
       The ``Plotter`` class **must** implement the ``get_dashboard(self)`` method.
       The ``get_dashboard`` method for this example case:
 
+      - Gets called once when the visualization actor is initialized.
       - Uses `HoloViews <https://holoviews.org/>`_ as its cornerstone to enable interactive visualizations.
       - Returns a `HoloViews DynamicMap <https://holoviews.org/reference/containers/bokeh/DynamicMap.html>`_ object, 
         which allows you to dynamically update a plot whenever its argument function is called, here ``self.plot_ip_vs_time``.
@@ -210,13 +212,13 @@ Exercise 1c: Creating a custom visualization
          the profile points in addition to time.
 
       Also implement the ``plot_f_df_dpsi_profile`` method in the ``Plotter`` class that 
-      displays the ff' profile as a function of psi for the current time step. 
+      displays the ff' profile stored in the state object as a function of psi for the current time step. 
 
       Your ``plot_f_df_dpsi_profile`` should do the following:
 
       - Load the state data from the current ``self.active_state``.
       - Extract the arrays for ff' and psi from the state data (use ``state.sel(time=self.time)``).
-      - Display psi on the x-axis and f_df_dpsi on the y-axis.
+      - Display psi on the x-axis and f_df_dpsi on the y-axis, using a `HoloViews Curve <https://holoviews.org/reference/elements/bokeh/Curve.html>`_.
       - Give an appropriate title, xlabel, and ylabel.
       - Properly handle the case when no data is available yet (Return an empty ``hv.Curve``).
 
@@ -293,11 +295,13 @@ Exercise 1c: Creating a custom visualization
 
       .. figure:: ../source/images/ff_prime.gif
 
+
 .. tip:: More complex examples of visualizations are available in the 
    ``imas_muscle3/visualization/examples/`` directory. For example, the PDS example
    combines data from multiple IDSs, handles machine description data, and 
    handles different types of plots.
 
+
 Exercise 2: Using Automatic Mode
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
@@ -320,6 +324,7 @@ Exercise 2: Using Automatic Mode
 
       - No fine grain control over the plots
       - Unable to combine data
+      - Slower performance and increased memory usage
 
       Repeat exercise 1a, however this time add the following settings to the yMMSL:
 
@@ -370,3 +375,104 @@ Exercise 3: Using the CLI
              "imas:hdf5?path=/home/ITER/blokhus/public/imasdb/ITER/4/666666/3/" \
              imas_muscle3/visualization/examples/simple_1d_plot/simple_1d_plot.py
 
+
+Exercise 4: Loading Machine Descriptions
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+.. md-tab-set::
+   .. md-tab-item:: Exercise
+
+      In this exercise you will create a 2D plot that combines **static machine description data** with **time-evolving equilibrium data**. Specifically,
+
+      - The plasma boundary outline (from the `equilibrium` IDS) as it changes over time.
+      - The tokamak first wall and divertor (from the `wall` machine description IDS).
+
+      You will need to update your yMMSL from exercise 1a, with the following changes:
+
+      - Add a new source actor which will send the ``wall`` and ``pf_active``
+        machine description IDSs to the visualization actor.
+      - The new source actor should use the following URI:
+         .. code-block:: bash
+
+            imas:hdf5?path=/home/ITER/blokhus/public/imasdb/ITER/4/666666/3/
+      - The visulization actor should receive the machine description IDSs on the S port, with the names ``wall_md_in`` and ``pf_active_md_in``.
+
+
+      You will need to implement the ``extract_equilibrium``, ``_plot_boundary_outline``, and ``_plot_wall`` methods, in the template below.
+
+      Implement the ``extract_equilibrium`` method which does the following:
+
+      - Loads R and Z coordinates of the the boundary outline from the equilibrium IDS: 
+        ``ids.time_slice[0].boundary.ouline.r``, ``ids.time_slice[0].boundary.ouline.z``
+      - Stores both in an Xarray Dataset.
+      - Either saves the first entry in ``self.data`` or concatenates it to an existing Dataset.
+
+      Implement the ``_plot_boundary_outline`` method which does the following:
+
+      - Load the state data from the current ``self.active_state``.
+      - Extract the r and z arrays from the state data (use ``state.sel(time=self.time)``).
+      - Display r and z, using a `HoloViews Curve <https://holoviews.org/reference/elements/bokeh/Curve.html>`_.
+
+      Implement the ``_plot_wall`` method which does the following:
+
+      - Load the wall machine description IDS using ``self.active_state.md.get("wall")``.
+      - Loads the first wall and divertor from the wall IDS:
+        ``ids.description_2d[0].limiter.unit[i].outline.r`` and ``ids.description_2d[0].limiter.unit[i].outline.z`` for ``i = 0`` and ``i = 1``.
+      - Display r and z, using a `HoloViews Path <https://holoviews.org/reference/elements/bokeh/Path.html>`_.
+
+      .. code-block:: python
+
+
+         import holoviews as hv
+         import numpy as np
+         import panel as pn
+         import xarray as xr
+
+         from imas_muscle3.visualization.base_plotter import BasePlotter
+         from imas_muscle3.visualization.base_state import BaseState
+
+
+         class State(BaseState):
+             def extract(self, ids):
+                 if ids.metadata.name == "equilibrium":
+                     self.extract_equilibrium(ids)
+
+             def extract_equilibrium(self, ids):
+                 # Implement this method!
+
+         class Plotter(BasePlotter):
+             DEFAULT_OPTS = hv.opts.Overlay(
+                 xlim=(0, 13),
+                 ylim=(-10, 10),
+                 title="Wall and equilibrium boundary outline",
+                 xlabel="r [m]",
+                 ylabel="z [m]",
+             )
+
+             def get_dashboard(self):
+                 elements = [
+                     hv.DynamicMap(self._plot_boundary_outline),
+                     hv.DynamicMap(self._plot_wall),
+                 ]
+                 overlay = hv.Overlay(elements).collate().opts(self.DEFAULT_OPTS)
+                 return pn.pane.HoloViews(overlay, width=800, height=1000)
+
+             @pn.depends("time")
+             def _plot_boundary_outline(self):
+                 # Implement this method!
+
+             def _plot_wall(self):
+                 # Implement this method!
+
+
+   .. md-tab-item:: Solution
+
+      Example yMMSL file:
+
+      .. literalinclude:: ../../imas_muscle3/visualization/examples/machine_description/machine_description.ymmsl
+
+      Example plotting script:
+
+      .. literalinclude:: ../../imas_muscle3/visualization/examples/machine_description/machine_description.py
+         :language: python
+

imas_muscle3/actors/visualization_component.py

@@ -60,7 +60,8 @@ def main() -> None:
     while instance.reuse_instance():
         if first_run:
             plot_file_path = instance.get_setting("plot_file_path", "str")
-            port = get_setting_optional(instance, "port", 5006)
+            # If port is not specified, use a random available port
+            port = get_setting_optional(instance, "port", 0)
             # FIXME: there is an issue when the plotting takes much longer
             # than it takes for data to arrive from the MUSCLE actor. As a
             # remedy, set a plotting throttle interval.

imas_muscle3/visualization/examples/machine_description/machine_description.py

@@ -0,0 +1,89 @@
+"""
+Example that plots the following:
+- First wall and divertor from machine description IDS.
+- Boundary outline from an equilibrium IDS
+"""
+
+import holoviews as hv
+import numpy as np
+import panel as pn
+import xarray as xr
+
+from imas_muscle3.visualization.base_plotter import BasePlotter
+from imas_muscle3.visualization.base_state import BaseState
+
+
+class State(BaseState):
+    def extract(self, ids):
+        if ids.metadata.name == "equilibrium":
+            self.extract_equilibrium(ids)
+
+    def extract_equilibrium(self, ids):
+        ts = ids.time_slice[0]
+        outline = ts.boundary.outline
+
+        boundary_data = xr.Dataset(
+            {
+                "r": (("time", "point"), [outline.r]),
+                "z": (("time", "point"), [outline.z]),
+            },
+            coords={
+                "time": [ids.time[0]],
+                "point": np.arange(len(outline.r)),
+            },
+        )
+
+        current_data = self.data.get("equilibrium")
+        if current_data is None:
+            self.data["equilibrium"] = boundary_data
+        else:
+            self.data["equilibrium"] = xr.concat(
+                [current_data, boundary_data], dim="time", join="outer"
+            )
+
+
+class Plotter(BasePlotter):
+    DEFAULT_OPTS = hv.opts.Overlay(
+        xlim=(0, 13),
+        ylim=(-10, 10),
+        title="Wall and equilibrium boundary outline",
+        xlabel="r [m]",
+        ylabel="z [m]",
+    )
+
+    def get_dashboard(self):
+        elements = [
+            hv.DynamicMap(self._plot_boundary_outline),
+            hv.DynamicMap(self._plot_wall),
+        ]
+        overlay = hv.Overlay(elements).collate().opts(self.DEFAULT_OPTS)
+        return pn.pane.HoloViews(overlay, width=800, height=1000)
+
+    @pn.depends("time")
+    def _plot_boundary_outline(self):
+        state = self.active_state.data.get("equilibrium")
+
+        if state is not None and "r" in state and "z" in state:
+            selected_data = state.sel(time=self.time)
+            r = selected_data.r.values
+            z = selected_data.z.values
+        else:
+            r, z = [], [], "Waiting for data..."
+
+        return hv.Curve((r, z)).opts(self.DEFAULT_OPTS)
+
+    def _plot_wall(self):
+        """Generates path for limiter and divertor."""
+        paths = []
+        wall = self.active_state.md.get("wall")
+        if wall is not None:
+            for unit in wall.description_2d[0].limiter.unit:
+                name = str(unit.name)
+                r_vals = unit.outline.r
+                z_vals = unit.outline.z
+                paths.append((r_vals, z_vals, name))
+        return hv.Path(paths, vdims=["name"]).opts(
+            color="black",
+            line_width=2,
+            hover_tooltips=[("", "@name")],
+        )

imas_muscle3/visualization/examples/machine_description/machine_description.ymmsl

@@ -0,0 +1,39 @@
+ymmsl_version: v0.1
+model:
+  name: test_model
+  components:
+    source_component:
+      implementation: source_component
+      ports:
+        o_i: [equilibrium_out]
+    source_component_md:
+      implementation: source_component
+      ports:
+        o_i: [wall_out, pf_active_out]
+    visualization_component:
+      implementation: visualization_component
+      ports:
+        s: [equilibrium_in, pf_active_md_in, wall_md_in]
+  conduits:
+    source_component.equilibrium_out: visualization_component.equilibrium_in
+    source_component_md.wall_out: visualization_component.wall_md_in
+    source_component_md.pf_active_out: visualization_component.pf_active_md_in
+settings:
+  source_component.source_uri: imas:hdf5?path=/home/ITER/blokhus/public/imasdb/ITER/4/666666/3/
+  source_component_md.source_uri: imas:hdf5?path=/home/ITER/blokhus/public/imasdb/ITER/4/666666/3/
+  visualization_component.plot_file_path: /home/ITER/blokhus/projects/IMAS-MUSCLE3/imas_muscle3/visualization/examples/machine_description/machine_description.py
+  visualization_component.keep_alive: true
+implementations:
+  visualization_component:
+    executable: python
+    args: -u -m imas_muscle3.actors.visualization_component
+  source_component:
+    executable: python
+    args: -u -m imas_muscle3.actors.source_component
+resources:
+  source_component:
+    threads: 1
+  source_component_md:
+    threads: 1
+  visualization_component:
+    threads: 1

imas_muscle3/visualization/examples/pds/pds.ymmsl

@@ -23,7 +23,6 @@ settings:
   source_component.source_uri: imas:hdf5?path=/home/sebbe/projects/iter_python/IMAS-MUSCLE3/data/pds_example
   source_component_md.source_uri: imas:hdf5?path=/home/sebbe/projects/iter_python/Waveform-Editor/data/nice-input-dd4
   visualization_component.plot_file_path: /home/sebbe/projects/iter_python/IMAS-MUSCLE3/imas_muscle3/visualization/examples/pds/pds.py
-  visualization_component.port: 5007
   visualization_component.keep_alive: true
 implementations:
   visualization_component:

imas_muscle3/visualization/examples/simple_1d_plot/simple_1d_plot.ymmsl

@@ -13,9 +13,8 @@ model:
   conduits:
     source_component.equilibrium_out: visualization_component.equilibrium_in
 settings:
-  source_component.source_uri: imas:hdf5?path=/home/sebbe/projects/iter_python/IMAS-MUSCLE3/data/striped_eq
-  visualization_component.plot_file_path: /home/sebbe/projects/iter_python/IMAS-MUSCLE3/imas_muscle3/visualization/examples/simple_1d_plot/simple_1d_plot.py
-  visualization_component.port: 5007
+  source_component.source_uri: imas:hdf5?path=/home/ITER/blokhus/public/imasdb/ITER/4/666666/3/
+  visualization_component.plot_file_path: /home/ITER/blokhus/projects/IMAS-MUSCLE3/imas_muscle3/visualization/examples/simple_1d_plot/simple_1d_plot.py
   visualization_component.throttle_interval: 0
   visualization_component.keep_alive: true
 implementations:

imas_muscle3/visualization/visualization_actor.py

@@ -1,6 +1,5 @@
 import logging
 import runpy
-import webbrowser
 from typing import Dict
 
 import panel as pn
@@ -92,18 +91,7 @@ def _start_server(self) -> None:
         self.server = pn.serve(  # type: ignore[no-untyped-call]
             self.dynamic_panel,
             port=self.port,
-            show=False,
+            show=self.open_browser_on_start,
             threaded=True,
             start=True,
         )
-        if self.open_browser_on_start:
-            self._open_browser()
-
-    def _open_browser(self) -> None:
-        """Open the dashboard in the system web browser."""
-        url = f"http://localhost:{self.port}"
-        try:
-            webbrowser.open(url)
-        except Exception as e:
-            logger.warning(f"Could not open browser automatically: {e}")
-        logger.info(f"Dashboard is available at {url}")